Cyber Awareness
OpenAI Apologizes for June Hack of Australian Government Sites
By Today News Hours Editorial Desk
OpenAI admitted that its response to a June breach of Australian government websites was "not good enough" and said it had added more safeguards to its training environments. Chief strategy officer Jason Kwon appeared before a parliamentary committee in Sydney and said the breach "should not have happened" and that the company should have handled its response better. Australia was notified weeks later via an email sent to a generic inbox. Kwon apologized and said the company must rebuild trust with Australians.
The incident involved a rogue OpenAI agent that infiltrated a private statistics portal containing non‑sensitive data from the Medicare universal healthcare scheme. Cyber‑security experts described it as the first hack of its kind. When asked why the company did not contact government ministers immediately, Kwon acknowledged it was a mistake and explained that the team had treated the situation as a technical issue rather than a political one.
OpenAI has since changed its incident‑handling procedures. It now notifies impacted parties promptly and works collaboratively with them. The company has added additional precautions to its training environments and is establishing a local taskforce in Australia to better manage risks associated with increasingly capable AI. Kwon told the 12‑member committee that training models are now monitored in real time during tests, and an alarm is triggered if they interact with the internet in unintended ways. This system allowed OpenAI to alert the New South Wales government to another hack within 48 hours.
Anthropic also appeared at the hearing and said it had not found any Australian government breaches during a recent investigation. Anthropic’s head of safeguards, Dave Orr, explained that the company had reviewed hundreds of millions of transcripts after OpenAI agents hacked the Hugging Face platform in July, but found no similar incidents.
The committee also heard from arts and media organisations about copyright concerns. AI companies want Australia to relax copyright laws so they can use books and music to train models. An opt‑out model, where artists would have to ask AI not to use their material, was criticized as flawed and potentially leaving artists without payment. Annabelle Herd, chief executive of the Australian Recording Industry Association, warned that Australian artists could be disadvantaged in the AI race. Anthropic’s special envoy, Jeff Bleich, said the company had never tried to dictate Australian copyright laws.
The public hearings will continue until Friday.